EP1876567A1 - Procédé de détermination du comportement dépendent du temps des objets mobiles et non-rigides, particulièrement des tissus biologiques, des données d'imagerie ultrasonique échographiques - Google Patents

Procédé de détermination du comportement dépendent du temps des objets mobiles et non-rigides, particulièrement des tissus biologiques, des données d'imagerie ultrasonique échographiques Download PDF

Info

Publication number
EP1876567A1
EP1876567A1 EP06425462A EP06425462A EP1876567A1 EP 1876567 A1 EP1876567 A1 EP 1876567A1 EP 06425462 A EP06425462 A EP 06425462A EP 06425462 A EP06425462 A EP 06425462A EP 1876567 A1 EP1876567 A1 EP 1876567A1
Authority
EP
European Patent Office
Prior art keywords
border
sequence
image
image frames
border line
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
EP06425462A
Other languages
German (de)
English (en)
Inventor
Stefano Pedri
Francesco Lo Nero
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Esaote SpA
Original Assignee
Esaote SpA
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Esaote SpA filed Critical Esaote SpA
Priority to EP06425462A priority Critical patent/EP1876567A1/fr
Publication of EP1876567A1 publication Critical patent/EP1876567A1/fr
Withdrawn legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/20—Analysis of motion
    • G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
    • G06T7/251—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments involving models
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/12—Edge-based segmentation
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
    • G01S7/52—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S15/00
    • G01S7/52017—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S15/00 particularly adapted to short-range imaging
    • G01S7/52053—Display arrangements
    • G01S7/52057—Cathode ray tube displays
    • G01S7/52073—Production of cursor lines, markers or indicia by electronic means
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10132—Ultrasound image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20092—Interactive image processing based on input by user
    • G06T2207/20096—Interactive definition of curve of interest
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20112—Image segmentation details
    • G06T2207/20168—Radial search
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30048—Heart; Cardiac

Definitions

  • the invention relates to a method of determining the time dependent behavior of non rigid moving objects, particularly of biological tissues from echographic ultrasound imaging data comprising the steps of tracking position and velocity of objects' borders in two or three dimensional digital images, particularly in echographic images comprising the following steps:
  • borders are drawn manually by the operator over the physiologically relevant frames of a sequence of images.
  • strain rate analysis method in ultrasonic diagnostic imaging is disclosed in WO 02/45587 .
  • strain rate analysis is performed for ultrasonic images in which the spatial gradient of velocity is calculated in the direction of tissue motion.
  • Strain rate is calculated for cardiac ultrasound images in the direction of motion which, for myocardial images, may be either in the plane of the myocardium or across the myocardium.
  • Strain rate information is calculated for a sequence of images of a heart cycle and displayed for an automatically drawn border such as the endocardial border over the full heart cycle.
  • the spatial gradient of velocity used for determining the strain and the displacements of the borders form one frame to a successive one in a sequence of frames uses so called Doppler Tissue Imaging, so called DTI or TDI.
  • This technique allows to measure tissue velocity over all points in the ventricular wall.
  • the measurement of velocity itself provides a direct information about the wall motion and helps to uncover abnormalities not immediately observable from the visualization in B-mode.
  • the velocity contains information about either rigid body displacement, shear, and contraction/distension, the latter being immediately related to the myocardial activity.
  • Post processing of the DTI velocity data allows the evaluation of additional quantities, namely strain-rate and strain, that are strictly related to the regional function. Segmental strain gives a direct evaluation of the degree of contractility of the myocardium during systole, as well as of its relaxation during ventricular filling.
  • TDI suffers from a few drawbacks consisting in limitations of the technique.
  • a Doppler signal requires additional processing with respect to the simple echo.
  • Doppler tissue imaging suffers further of an intrinsic limitation due to the fact that only the component of velocity along a scanline can be measured. This limitation has several drawbacks. When tissue moves in a direction that is not aligned with the scanline, the Doppler velocity does not reflect the effective tissue kinematics. Only the component of strain and strain rate along the scanline can be evaluated correctly, giving a reduced view of the local deformation state. This limits the application of DTI (or TDI) to the anatomic sites that can be imagined aligned along a scanline. In echocardiography this corresponds essentially to the interventricular septum and to the lateral walls in apical view.
  • DTI or TDI
  • An object of the present invention is to provide for such a method of determining the time dependent behavior of non rigid moving objects, particularly of biological tissues from echographic ultrasound imaging data which is independent from the angle of propagation of the ultrasound beams and which method makes use of a method of tracking position and velocity of objects' borders in two or three dimensional digital images, particularly in echographic images, elaborating the said ultrasound imaging data in order to track the displacement of a border of a moving tissue or a moving object in a sequence of consecutive ultrasound image frames.
  • Another object of the present invention is to provide for a method of displaying the results of the dynamical behavior of the moving object, particularly of the moving tissue in combination of a two or three dimensional image of the object and of its surroundings. More specifically the moving object is the ventricular wall of the left ventricle and the dynamical behavior of the said ventricle wall is displayed overlaid or combined with the B-mode image of the said ventricle.
  • the present invention achieves the above mentioned aims with a method of determining the time dependent behavior of non rigid moving objects, particularly of biological tissues from echographic ultrasound imaging data comprising the steps of tracking position and velocity of objects' borders in two or three dimensional digital images, particularly in echographic images comprising the following steps:
  • the tangential component of the velocity vector at the selected points on the border line chosen are a measure for a twisting motion of the ventricular wall, while the radial component of the velocity vector at the said selected points are a measure of a deformation of the ventricle wall in the sense of the constriction of the ventricle.
  • the sequence of image frames acquired is a sequence of consecutive B-mode, grey scale ultrasound images.
  • the method provides the following steps:
  • the original trace of pixels coinciding with the manually or automatically drawn border line is followed in time, i.e. in the at least one following image frame by searching the maximum likelihood of the trace of pixels in the following image frame with the trace of pixels in the first or timely previous image frame of the sequence of image frames by analyzing the image pixels in the neighborhood of the said trace of pixels.
  • the tracking of the border line is carried out by defining a certain number of reference points on the manually or automatically drawn border line on the first image frame and by using the method of the so called transmural cuts.
  • the said method of transmural cuts is disclosed in greater detail in the document PCT/IT02/00114 filed on 27.02.2002.
  • More precisely transmural cuts consist in defining a line which crosses the border line drawn and passing through one reference point.
  • a physiologically appropriate direction can be chosen, which typically can be the orthogonal direction to the border line at the reference point.
  • This operation is made for each image frame of the sequence of frames and for each reference point chosen.
  • the pixels taken along each transmural line in each of the image frames of the sequence of image frames are placed in columns, each column corresponding to one frame of the sequence of images. In this way the evolution along a transmural cut, can be represented for all instants at once in a two-dimensional space time representation.
  • the above disclosed procedure is a reduction of a two dimensional problem applied to a two dimensional image such as a B-mode ultrasound image to a one dimensional problem as a M-mode image.
  • the tracking of the border. i.e. of the trace of pixels is carried out along the space-time image using a cross-correlation procedure of the pixel column in the space-time image corresponding to a first image frame with the pixel column in the space-time image corresponding to a successive image frame of the sequence of image frames.
  • This technique can be applied to any kind of images in which the geometry of the border line drawn does not require any kind of special reference points to be tracked a priori such as pro example closed border lines as the border line of the cavity of a blood vessel in a cross-section image of the vessel.
  • a preventive cycle must be carried out for optimally tracking the border-line of the object along the sequence of image frames.
  • the general topology of the border line of the object imaged can be represented by tracking the motion of these few representative points prior to carry out the tracking of at least one or some of the reference points lying on the manually or automatically drawn border-line in each frame of the sequence of image frames.
  • These representative points can be for example the starting and ending points of the border line when this is an open one.
  • the representative reference points of the border of an imaged object can be also suggested by the physiology when the imaged object is a particular tissue or organ, such as for example the left ventricle.
  • This few representative reference points is carried out in a identical way as the one disclosed above for the other reference points on the border-line drawn manually or automatically on the first frame by using the method of transmural cuts for constructing space-time images of each of the few representative reference points and determining the displacement of these points in each of the frames of the sequence of image frames by means of cross-correlation between each of the pixel columns with the successive pixel column corresponding to the pixels along the transmural cut across the same representative point in the different image frames of the sequence of image frames.
  • the direction of the transmural cuts can be chosen as the orthogonal direction to the border line at the corresponding reference point.
  • the position and the displacement of the other reference points on the border-lines at each image frame of the sequence of image frames are obtained by rescaling the originally drawn border-line in the first image frame in such a way to obtain in each image frame corresponding to a successive instant a topologically equivalent border line geometry with respect to the original border line. Typically this results in a translation of all points along the original border line.
  • This preliminary rescaling allows to keep the representative reference points always in the proper position in all frames of the sequence of image frames by rearranging the other reference points so that the representative reference points maintains the same meaning relatively to the object in all frames of the sequence of image frames.
  • the space-time representation along the transmural cuts can be built using a line for the transmural cut with a thickness larger than that of a single pixel and by extracting the average value across such a thickness.
  • the above mentioned method can be further developed for carrying out a surface border tracking three dimensional imaging.
  • the method according to the said development comprises the following steps:
  • one or more further principal section planes can be defined along each of which further section planes the methods steps 1) to o) are carried out.
  • two orthogonal principal section planes are chosen for carrying out the above mentioned method steps, the crossing line of the two principal section planes defining a preferred direction of the said planes.
  • the said direction can be chosen as suggested by the topological or functional feature of the object imaged.
  • this physiologically relevant direction can be chosen as the cut across a central vertical plane such as the ventricle axis.
  • the method according to the present invention comprises the steps of defining bounds or limits for a distance range within which the share of the said secondary section planes is defined.
  • a topological or physiological relevant direction is chosen, particularly the same direction defined for determining the principal section planes, along which direction bounds are determined for the ends of a distance range within which the share of secondary section planes at least transversal, particularly perpendicular to the said relevant direction is determined.
  • this one is determined as a physiologically relevant line passing through the reliable points.
  • the correct border is determined along a sequence of two dimensional or three dimensional ultrasound image data and the correct border for each image frame can be displayed overlaid on the displayed image frame as an highlighted line characterized by a color which is different from the grey-scale B-mode image displayed.
  • the component(s) of velocity in the direction of the transmural cut(s) can be estimated for each reference point by means of a simple calculation.
  • the complete velocity vector is determined by evaluation of the other component(s) of velocity, the total number of components being two for two-dimensional imaging and being three for three-dimensional imaging.
  • a transmural cut consisting in a line which crosses the tracked point and directed along the direction where the additional component of velocity must be evaluated, typically orthogonal to the direction previous employed for tracking the border.
  • the evaluation of the velocity component along the chosen direction is carried out along the space-time image using a cross-correlation procedure of the pixel column in the space-time image.
  • the velocity being given by the ratio of the column-wise displacement of the correlation maximum and the time interval between the corresponding frames.
  • the method is identical of that employed for tracking the border with the difference that only the frame-by-frame displacement is required and the eventual time integration of said displacement to get the motion of the border is ignored.
  • a different evaluation of the velocity vector can be obtained by applying two dimensional-correlation techniques or a specific optical flow technique particularly developed for ultrasound image data of moving objects.
  • the said velocity estimation method can be carried out in combination with the above disclosed method for tracking the border of the imaged moving object.
  • the said method is an adaptation of known method so called OPTICAL FLOW methods, like a known method so called PIV method used in fluid dynamics.
  • the border tracked can be drawn as a line as disclosed above and the velocity vectors of the border taken at certain number of points of the said border line are displayed as arrows having a different color as the border line and the direction of the velocity vector and a length corresponding to the modulus of the velocity vector in the image plane of the two dimensional image displayed.
  • the border line tracked is displayed as a line and the instant velocity vectors also, the line and the vectors being drawn as color highlighted line on the grey scale B-mode image.
  • the invention can provide alternative view in which only the tangential and only the radial component of the velocity vectors are shown or the said velocity vectors are shown.
  • one image can show superimposed or drawn on the grey scale B-mode image the border line, the velocity vector and one or both components oriented in the tangential direction and/or in the radial direction.
  • tracking technique can be used as for example the so called PIV or the optical flow technique used in so called registering methods which are described in detail in the reference already cited above.
  • the automatic tracking method disclosed here allows the tracking of a border on a sequence of two-dimensional or three-dimensional images, and the evaluation of the velocity vector field on such borders.
  • the border could be tracked on the basis of the velocity vector only, however a tracking procedure is a result of the summation (time integration) of the estimated velocities and is prone to an error growth in presence of small incorrect estimates.
  • This approach reduces the two- or three-dimensional tracking to a combination of one-dimensional tracking problems along the single topological relevant direction (typically the orthogonal to the border), that can be much better controlled and made accurate.
  • the accurate tracking result is employed to improve the estimates of the velocity vector.
  • the result of this procedure is the automatic definition of the borders displacement and velocity over all frames of a sequence of images, starting from the border traced on a single image.
  • the found borders information will be used to evaluated some geometric properties, like volume, area, or sizes, of the organ.
  • the border kinematics tilt + velocity
  • the velocity vectors having absolute modulus and direction are obtained without any angular dependence from the direction of propagation of the ultrasound beams relatively to the imaged object. This allows to determine the components or projections of the said velocity vectors onto two direction which are perpendicular one to the other and one of which is oriented tangentially to the border line at the corresponding reference point. In this condition velocity information in the direction of the border line and thus of the tissue wall and in the direction normal to the said tissue wall can be determined at each reference point which velocity data are a measure of the strain or strain rate in the direction of the tissue wall and in a direction perpendicular to it.
  • the present invention allows to obtain velocity information along the circumferential direction of the annular shape of the ventricle wall as seen in the said short axis view and in the radial direction of the said annular shape of the ventricle wall in the said short axis view.
  • Circumferential strain and strain rate are measures of the twisting of the ventricle wall around the longitudinal axis, which is an axis going form the apex to the mitral valve. Radial strain and strain rate give information about the constriction or compression of the ventricle.
  • the method steps according to the present invention are firstly described with reference to a two dimensional case.
  • a sequence of two dimensional B-mode image frames is acquired.
  • the frames are acquired at predetermined time interval one form the other.
  • the images contain one organ/object or part of it, that changes its position and shape in time, of which organ we want to determine the dynamical behavior during motion by trace the border kinematics at all instants.
  • a typical organ for which the above conditions are met is the heart. Echocardiographic investigations have a great importance. Strain rate and strain echocardiography is an emerging non invasive technique for assessing myocardial function. Regional myocardial strain rate and strain can detect inducible ischemia at earlier stages than visual estimation of wall motion or wall thickening parameters. Changes in systolic and diastolic strain rate and strain have the potential to discriminate between different myocardial viability states.
  • Figure 1 is a diagrammatic representation of the heart. Here the left ventricle and the right ventricle are indicated and also the apex of the left ventricle.
  • the three lines indicates three section planes or slices along which so called short-axis views of the heart are taken.
  • the images taken along a certain slice are of the so called B-mode echographic image type.
  • Figure 2 shows diagrammatically the image which should be seen along a short axis view slice corresponding to the central line in figure 1 and which corresponds to the so called mid-ventricular short axis view.
  • Longitudinal or apical strain and strain rate gives information about the longitudinal mechanics of the heart as a pump essentially at the Left ventricular annulus and at the Mid and Basal segments of the septum. The direction of the strain or strain rate vectors are indicated in figure 3A.
  • Figure 3B indicates the radial strain or strain rate and figure 3B the circumferential strain or strain rate vectors as important parameters giving information on the radial and circumferential mechanics of the left ventricle.
  • lengthening is represented as a positive value for strain, while shortening is represented by a negative value. Being the initial instant t0 taken at the end of diastole, Strain usually assumes negative values.
  • SR Stress Strain Rate
  • the measurement unit of SR is [1/s] or [s-1]
  • the present invention is based on a new angle independent technique which allows to determine the velocity of every point on a line drawn to represent an imaged wall or border of a moving object and particularly of a moving organ such as the heart.
  • This technique provides for a border tracking to estimate tissue velocity at a collection of point in two-dimensional or three-dimensional sequence (loop) of B-Mode images.
  • the Myocardial borders are followed over the time starting from one reliable existing instantaneous trace or border line, either drawn by the physician over one single frame or determined by automatic border detection algorithm.
  • the velocity is displayed as a vector overlaid onto the B-mode image where the length of the vector indicates the magnitude of the tissue velocity representation.
  • Figure 4 illustrates an example of such results in a short axis view of the left ventricle.
  • a short axis B-mode image of the left ventricle is shown.
  • the same short-axis B-mode image as in the upper half is shown but on it there is superimposed the border line and the instant velocity vectors at certain reference points on the said borderline.
  • the border line and/or at least the reference points on it are tracked over every frame of the sequence and the velocity vector at each reference point is determined and displayed as a function of the difference in position of the reference point in a preceding and in a following frame and of the time interval between the these frames.
  • the velocity vectors on the screen may vary with time if a sort of cinema view is carried out.
  • the velocity vectors determined as described above contain information about the dynamical behavior of the object, in this case the left ventricle wall.
  • this information In order to be able to use this information for correlating mechanical behavior to physiological meaning so that only diagnostically relevant information is furnished to the specialized person, it is useful to isolate the information relatively to the tangential and to the radial strain and strain rates as explained and defined above.
  • Figures 5 and 6 illustrates the results of the steps according to the present invention which provides for determining for each of the reference points on the border line the projection of the velocity vector determined at the said reference point onto the tangential direction and onto a direction which is normal to the tangential direction.
  • tangential direction it is meant the direction which is the one of the tangent line to the border line at the corresponding reference point.
  • radial direction it is meant the direction of a line which is normal to the said tangent line and which in a circle corresponds to the direction of the radius passing through the corresponding reference point.
  • the upper half part of figure 5 illustrates the same short-axis B-mode image of the left ventricle as in figure 4.
  • the lower half the said B-mode image is the same B-mode image as in the upper half but on it there is superimposed the border line with the reference points and for each reference point on the said border line the tangential component at the said reference points of the instant velocity vectors represented on figure 4.
  • the tangential component is represented as an arrow oriented in the direction of the tangent to the border line at the corresponding reference point and the length of the arrow indicates the velocity in the said direction and at the corresponding reference point.
  • figure 6 shows in the upper half the B-mode image illustrated in figures 4 and 5 while in the lower half on these B-mode image there are also displayed the border line and the reference points on it while the arrows indicates the radial component of the instant velocity vector at each of the reference points.
  • Each arrow indicates the radial direction at the corresponding reference point and the length of each arrow indicates the modulus of the said velocity component in the radial direction at each reference point of the border line.
  • the two velocity components i.e. the tangential and the radial component of the instant velocity vectors do not represent directly the strain and the strain rate, it is immediately clear to the skilled person that differences in the modulus of the tangential direction among the different reference points is a measure of the deformation of the border line and thus of the ventricle wall represented by the said border line. From the tangential and from the radial components of the velocity vectors at different points on the border line it is a straight mathematical way to determine the strain and the strain rate either in circumferential and in radial direction of the border line and thus of the ventricle wall in its short axis view.
  • the meaning of the circumferential and radial strain or strain rate becomes clear.
  • the measure of a twisting of the ventricle wall around a longitudinal or apical axis and of the change of this twisting in time is determined.
  • the radial strain and strain rate is a measure of the compression of the ventricle wall in the sense od a constriction of the ventricle volume and of the change in this compression in time.
  • the method according to the present invention requires to determine the velocity vectors in an independent manner from the incidence angle of the echographic beam for generating the B-mode image of the moving object.
  • the present invention provides a special tracking method which is particularly advantageous in combination with the above described method for determining the strain or strain rates of moving objects, particularly of the walls of the said objects and in the special case of echo cardiology.
  • a border is traced, manually or by another manual or automatic procedure, over one arbitrary frame.
  • Figure 7 illustrates an image of the left ventricle where the endocardial border points are traced from one side of the mitral annulus to the other side of the same mitral annulus.
  • Figure 8 illustrates an example in which the border is a closed one where the Nth point connects to the first one.
  • the method according to the invention provides for a second step of tracking the most representative reference point of the border line drawn in the first image frame.
  • the general topology of the object border is reproduced on all the images by tracking the motion of a few representative points. These are commonly the starting and final points of the border when this is an open one. In specific cases additional reference points can be added to improve the first evaluation of the region about which the border must be sought.
  • Figure 9 shows the reference points for a left ventricle (in long axis view) that are the starting and final points of the originally traced border.
  • the physiology suggests to track the motion of these points in the direction instantaneously orthogonal to the mitral plane (that is defined by these points).
  • the tracking along a specified direction is performed by using the method of transmural cuts as follow.
  • a line crossing the wall, passing through the point, and directed along the physiologically appropriate direction is drawn; in the case shown in figure 9 the appropriate direction is orthogonal to the mitral plane.
  • two orthogonal direction can be employed.
  • the pixels taken along the chosen direction line(s) are placed in columns, each column corresponding to one frame of the sequence of images.
  • the evolution along a line can be represented for all instants at once in a two dimensional space-time representation (sometime referred as M-mode representation) where one axis is the distance along the line and the other axis is the time.
  • M-mode representation two dimensional space-time representation
  • the space time representation can be built using a line for the transmural cut with a thickness larger than that of a single pixel and extracting the average value across such a thickness. The border tracking is then performed along the space-time image.
  • the tracking procedure is a procedure for following a border along one direction in a two-dimensional image like that in figure 4 starting from a known position at one instant.
  • the displacement from the known point yk to the point yk+1 can be estimated by evaluating the cross-correlation between the entire column at xk with the entire column at xk+1.
  • the first estimate is improved by applying the same procedure recursively on increasingly reduced spatial width about the previously found border.
  • This first estimate yi can be further improved.
  • a subset of the image is extracted by taking a few points above and below the first estimate yi, and a new image whose center corresponds to the sequence yi is generated.
  • a snake procedure like the one described in Blake A., Yuille A. Active Vision MIT press, 1992 ., is employed to follow, in the new image, the image brightness level that passes through the fixed point yk.
  • the tracking technique is a unique procedure that is common to different steps of the method according to the present invention.
  • the result of this preliminary tracking procedure is the position and displacement, at all instants, of the most representative reference points along the predefined direction, or the vector combination when two directions are employed.
  • This preliminary rescaling procedure permits to keep the reference points always at the proper position in all the frames, and to rearrange the other points so that the reference maintains the same meaning in all the frames.
  • the present step of tracking the most representative reference points such as the starting and ending point of a border line can be avoided when the specific geometry does not require or have any representative reference point to be tracked a priori.
  • One example where this step can be avoided is given by the closed geometry in figure 8.
  • the method according to the invention provides for a further step consisting in the tracking of all the other reference point on the border line drawn manually or automatically in the first step on a first two dimensional image frame of the sequence of image frames.
  • the tracking along a specified direction is performed by using the method of transmural cuts as follow.
  • a line crossing the wall, passing through the point, and directed along the physiologically appropriate direction is drawn, this operation is made for each instant/frame of the sequence of image frames because the points are not fixed in time but they have been previously rescaled at each instant accordingly with the instantaneous displacement of the reference points.
  • the appropriate direction is taken at each instant as orthogonal to the rescaled border.
  • the pixels taken along each transmural line are placed in columns, each column corresponding to one frame of the sequence of images. In this way the evolution along a transmural cut, that is not fixed in all frames time but is slightly modified accordingly to the rescaling, can be represented for all instants at once in a two-dimensional space time representation analogous to that shown in figure 10.
  • the space time representation can be built using a line for the transmural cut with a thickness larger than that of a single pixel and extracting the average value across such a thickness.
  • the border tracking is then performed along the space-time image using the same technique employed in the step of tracking the representative reference points and disclosed above in a detailed manner.
  • the result of this step is the position, at all instants, of all the points along the predefined direction, or the vector combination when two directions are employed.
  • all the original points have been tracked in time, each one independently, and we have a new border tracked over all frames.
  • the method according to the present invention can be provided in combination with a procedure for determining the instant border line velocity vector for each one of the reference pints defined on the border line as tracked on each two dimensional frame.
  • the velocity vector can be known when two direction (three for three-dimensional imaging) are employed for displacing it.
  • the complete velocity vector can be evaluated by selecting additional direction for the transmural cuts on the already displaced point and evaluating the velocity along the additional direction.
  • the space time representation can be built using a line for the transmural cut with a thickness larger than that of a single pixel and extracting the average value across such a thickness.
  • the complete velocity vector can be evaluated by a two-dimensional correlation technique or a specific optical flow technique adapted to the particular case of ultrasound imaging B-mode data.
  • the two-dimensional result can then be improved by imposing its accordance with the previous estimate obtained for one component from the transmural cut approach. Results of the entire procedure are shown, for one frame, in figure 7 and 8.
  • the method according to the present invention has been described until now as referred to only a two dimensional imaging. It is clear to the skilled person that the same method can be extended to the three dimensional case.
  • the geometrical definitions of tangent and radial or normal has the same meaning as in the two dimensional case only extended to the three-dimensional representation.
  • the main problem is the one of extending the tracking of the borderline in the three-dimensional case. This is described in greater detail in the following.
  • a sequence of three-dimensional (3D) datasets is mathematically a four-dimensional (4D) information that is 3D in space and 1D in time.
  • the images contain one organ/object or part of it, that changes its position and shape in time, of which organ it is desired to trace the border at all instants, the border now being a sequence of two-dimensional surfaces.
  • the method according to the present invention provides to choose one principal section plane which cuts to the three-dimensional dataset, and to apply the entire two-dimensional technique disclosed above on such plane.
  • the principal section plane of the 3D dataset is one plane, preferably along a physiologically relevant direction. Cutting the 3D datasets of the sequence of 3D datasets with this plane furnishes one sequence of 2D images.
  • Figure 12 illustrates the cutting of a three-dimensional data set of ultrasound image data of the object 0 with two orthogonal principal section planes 1 and 2 oriented in the vertical direction.
  • the above steps can be repeated with more than one or two principal section planes to improve the reliability of following steps in poor quality images.
  • a further step is carried out consisting in defining secondary section planes to the three-dimensional dataset, and applying the two-dimensional technique on single frames substituting the time direction with one spatial direction.
  • the previous step allows to define the bounds of the surface border. For this, one direction is chosen over the plane cut used in the previous step, preferably a physiologically relevant one (like the ventricle axis), and, for each instant, evaluate the upper and lower bounds along such direction of the border found in the previous step.
  • a reliable border in one single frame is defined, commonly the same frame used when the borders are drawn manually during the previous step relative to the principal section planes.
  • the border now contains one or more reliable points, at the intersection with the principal section plane or planes 1, 2 and that come from the border(s) determined in the previous step relative to the principal section planes as illustrated in fig. 14 and indicated by R1, R2, R3, R4.
  • a first guess border is constructed as a physiological relevant one passing through these reliable points R1, R2, R3, R4.
  • An example of the said guess border on a secondary section plane is illustrated in the example of figure 15.
  • the two dimensional image on a secondary section plane is illustrated together with the two reliable points R1 and R2.
  • the guess border passing through the said two reliable points R1 and R2 is given by given by a circle in the transversal images of the left ventricle.
  • a new border is now detected by the same procedure used for a single transmural cut as disclosed in the previous chapter for the two dimensional case, this time however, substituting the time coordinate with the spatial coordinate along such first guess border as follows. Make a number of transmural cuts on the single image along the guess border, place the pixel found along each cut side by side in a new two-dimensional image and obtain a new image, like that in figure 4, where the horizontal axis does not indicate the time coordinate but the spatial coordinate along the tentative border. As a result the correct border is tracked in one frame for each of the M sequences.
  • the instantaneous velocity vector for a certain number of predefined points on the border surface can be calculated by using the same technique disclosed of the two dimensional case.
  • the two dimensional technique disclosed above is used here by substituting the two dimensional estimate with a three-dimensional estimate of velocity.
  • this is done by using a three-dimensional correlation or optical flow technique, in place of the two-dimensional one for evaluating the three-dimensional velocity vector.
  • the corresponding B-mode image is displayed in a grey scale while the correct border line tracked is displayed overlaid on the corresponding displayed B-mode image of a corresponding image frame.
  • the said border line is in the form of an highlighted line characterized by a colour which is different from the grey-scale B-mode image displayed and on the border line the selected reference points tracked are displayed as highlighted dots with a colour which is different from the grey-scale B-mode image displayed, while the instant velocity vectors are displayed also superimposed in the grey scale B-mode image for each or part of the reference points, the direction of the arrow corresponding to the direction of the instant velocity vector and the lengths of the arrow being proportional to the modulus of the said instant velocity vector represented by the said arrow.
  • arrows displayed representing the tangential component and or arrows representing the radial or normal component of the said velocity vectors the direction of the arrows being parallel to the tangential or radial direction of the said tangential or radial components of the velocity vector and the length of the arrows being proportional to the modulus of the said tangential or radial component of the velocity vector.
  • the time dependent modulus of the modulus of the tangential component, of the radial component, of the velocity vector is displaied in form of a curve over time.
  • the two left lower areas display the time dependent strain and strain rate in the form of a curve over time.
  • FIG. 1 may represent similarly the tangential strain and/or the tangential strain rate, and/or the radial strain and/or the radial strain rate and/or the velocity vector modulus at each of the selected points.
  • Each selected point on the border line is displayed in a different colour univoquely associated to the said point, while the curve related to the time dependent velocity or strain or strain rate are displayed with the colour of the corresponding point in each of the above mentioned diagrams provided.
  • the said curves are displayed overlaid in the same diagram or coordinate system for each category of value represented, namely, velocity, velocity rate, strain, strain rate.
  • the curves of the time dependency of the direct distance or of the distance along the border line between respectively to adiacent points of the selected points on the border line are displayed. This is done similarly to the above disclosed cases by highlighting each of the said curves with a different colour which is the colour of one of the two points which distance is represented by the said curve.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • Ultra Sonic Daignosis Equipment (AREA)
EP06425462A 2006-07-04 2006-07-04 Procédé de détermination du comportement dépendent du temps des objets mobiles et non-rigides, particulièrement des tissus biologiques, des données d'imagerie ultrasonique échographiques Withdrawn EP1876567A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
EP06425462A EP1876567A1 (fr) 2006-07-04 2006-07-04 Procédé de détermination du comportement dépendent du temps des objets mobiles et non-rigides, particulièrement des tissus biologiques, des données d'imagerie ultrasonique échographiques

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
EP06425462A EP1876567A1 (fr) 2006-07-04 2006-07-04 Procédé de détermination du comportement dépendent du temps des objets mobiles et non-rigides, particulièrement des tissus biologiques, des données d'imagerie ultrasonique échographiques

Publications (1)

Publication Number Publication Date
EP1876567A1 true EP1876567A1 (fr) 2008-01-09

Family

ID=37478944

Family Applications (1)

Application Number Title Priority Date Filing Date
EP06425462A Withdrawn EP1876567A1 (fr) 2006-07-04 2006-07-04 Procédé de détermination du comportement dépendent du temps des objets mobiles et non-rigides, particulièrement des tissus biologiques, des données d'imagerie ultrasonique échographiques

Country Status (1)

Country Link
EP (1) EP1876567A1 (fr)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101926648B (zh) * 2009-08-03 2012-01-25 浙江工业大学 基于四维医学图像的心壁应力应变测量方法
JP2013022462A (ja) * 2011-07-19 2013-02-04 Toshiba Corp 運動対象輪郭トラッキング装置、心筋運動解析装置、運動対象輪郭トラッキング方法および心筋運動解析方法
CN112545564A (zh) * 2019-09-26 2021-03-26 通用电气公司 超声波诊断装置以及其控制程序

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040143189A1 (en) * 2003-01-16 2004-07-22 Peter Lysyansky Method and apparatus for quantitative myocardial assessment
EP1522875A1 (fr) * 2003-09-30 2005-04-13 Esaote S.p.A. Procédé pour suivre la position et la vitesse des contours d'un objet dans des images à deux ou trois dimensions, en particulier dans des images d'échographie

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040143189A1 (en) * 2003-01-16 2004-07-22 Peter Lysyansky Method and apparatus for quantitative myocardial assessment
EP1522875A1 (fr) * 2003-09-30 2005-04-13 Esaote S.p.A. Procédé pour suivre la position et la vitesse des contours d'un objet dans des images à deux ou trois dimensions, en particulier dans des images d'échographie

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
COOK A.: "Ultrasound Delivers New Insight into the Heart", MEDICAL SOLUTIONS, June 2006 (2006-06-01), pages 68 - 71, XP002410938, Retrieved from the Internet <URL:http://healthcare.siemens.com/magazine/documents/2006_04/p68-71_care_vvi_june_06_eng.pdf> [retrieved on 20061208] *
VANNAN ET AL.: "Effect of Cardiac Resynchronisation Therapy on Longitudinal and Circumferential Left Ventricular Mechanics by Velocity Vector Imaging: Description and Initial Clinical Application of a Novel Method Using High-Frame-Rate B-Mode Echocardiographic Images", ECHOCARDIOGRAPHY, vol. 22, no. 10, November 2005 (2005-11-01), pages 826 - 830, XP002410937 *

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101926648B (zh) * 2009-08-03 2012-01-25 浙江工业大学 基于四维医学图像的心壁应力应变测量方法
JP2013022462A (ja) * 2011-07-19 2013-02-04 Toshiba Corp 運動対象輪郭トラッキング装置、心筋運動解析装置、運動対象輪郭トラッキング方法および心筋運動解析方法
EP2549436A3 (fr) * 2011-07-19 2013-04-24 Kabushiki Kaisha Toshiba Appareil et procédé de suivi de contour d'objet mobile et appareil et procédé d'analyse de mouvement du myocarde
EP2738741A1 (fr) * 2011-07-19 2014-06-04 Kabushiki Kaisha Toshiba Appareil et procédé de suivi de contour d'objet mobile et appareil et procédé d'analyse de mouvement du myocarde
US9171220B2 (en) 2011-07-19 2015-10-27 Kabushiki Kaisha Toshiba Apparatus and method for tracking contour of moving object, and apparatus and method for analyzing myocardial motion
CN105105775A (zh) * 2011-07-19 2015-12-02 株式会社东芝 心肌运动解析装置
US9275266B2 (en) 2011-07-19 2016-03-01 Kabushiki Kaisha Toshiba Apparatus and method for tracking contour of moving object, and apparatus and method for analyzing myocardial motion
CN105105775B (zh) * 2011-07-19 2018-11-09 东芝医疗系统株式会社 心肌运动解析装置
CN112545564A (zh) * 2019-09-26 2021-03-26 通用电气公司 超声波诊断装置以及其控制程序
US11559286B2 (en) * 2019-09-26 2023-01-24 General Electric Company Ultrasound diagnostic apparatus and control program thereof for detecting the three dimensional size of a low echo region

Similar Documents

Publication Publication Date Title
EP1522875B1 (fr) Procédé pour suivre la position et la vitesse des contours d&#39;un objet dans des images digitales d&#39;échographie à deux ou trois dimensions.
EP0880937B1 (fr) Système d&#39;imagerie ultrasonique pour le diagnostic à poursuite Doppler du mouvement du tissu
US6503202B1 (en) Medical diagnostic ultrasound system and method for flow analysis
US20080095417A1 (en) Method for registering images of a sequence of images, particularly ultrasound diagnostic images
US11969286B2 (en) Systems and methods for automatic detection and visualization of turbulent blood flow using vector flow data
US6994673B2 (en) Method and apparatus for quantitative myocardial assessment
US5820561A (en) Analysis and measurement of temporal tissue velocity information
US6780155B2 (en) Method and system for ultrasound blood flow imaging and volume flow calculations
US7794398B2 (en) Real-time volumetric bi-plane ultrasound imaging and quantification
US8029444B2 (en) Method for estimating tissue velocity vectors and tissue deformation from ultrasonic diagnostic imaging data
US20030171668A1 (en) Image processing apparatus and ultrasonic diagnosis apparatus
Adam et al. Semiautomated Border Tracking of Cine Echocardiographic Ventnrcular Images
US20110313291A1 (en) Medical image processing device, medical image processing method, medical image diagnostic apparatus, operation method of medical image diagnostic apparatus, and medical image display method
EP4193928A1 (fr) Diagnostic ultrasonore de performance cardiaque utilisant une segmentation de chambre de modèle cardiaque avec commande d&#39;utilisateur
JP7232195B2 (ja) 血管内の壁せん断応力の同時視覚化及び定量化のためのシステム及び方法
US6909914B2 (en) Method for generating time independent images of moving objects
EP1772825A1 (fr) Méthode d&#39;alignement des images d&#39;une séquence d&#39;images, en particulier des images diagnostiques d&#39;ultrasons
EP1876567A1 (fr) Procédé de détermination du comportement dépendent du temps des objets mobiles et non-rigides, particulièrement des tissus biologiques, des données d&#39;imagerie ultrasonique échographiques
US20180049718A1 (en) Ultrasonic diagnosis of cardiac performance by single degree of freedom chamber segmentation
EP4210588A1 (fr) Systèmes et procédés de mesure de la rigidité cardiaque

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IS IT LI LT LU LV MC NL PL PT RO SE SI SK TR

AX Request for extension of the european patent

Extension state: AL BA HR MK YU

AKX Designation fees paid
REG Reference to a national code

Ref country code: DE

Ref legal event code: 8566

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE APPLICATION IS DEEMED TO BE WITHDRAWN

18D Application deemed to be withdrawn

Effective date: 20080710